3. Learning Objectives
• Understand the key differences between
supervised and unsupervised learning
• Learn about various use cases and real-world
applications
• Explore example datasets and case studies
4. What is Supervised Learning?
• Definition: Supervised learning is a type of machine
learning where the model is trained on labeled data.
• Key Components:
• Inputs (Features): Data attributes used for prediction
• Outputs (Labels): The correct answers provided in training
• Learning Process: Model learns by mapping inputs to
outputs
• Examples:
• Email Spam Detection (Spam or Not Spam)
• Credit Score Prediction
5. What is Unsupervised Learning?
• Definition: Unsupervised learning is a type of
machine learning where the model is trained on
unlabeled data.
• Key Components:
• No Labeled Output: The model identifies patterns and
structures
• Clustering & Association Rules: Finding hidden
relationships
• Examples:
• Customer Segmentation for Marketing
• Anomaly Detection in Fraud Detection
6. Key Differences – Supervised vs. Unsupervised
Learning
Feature Supervised Learning Unsupervised Learning
Data Type Labeled Data Unlabeled Data
Learning Process Maps input to known
output
Finds patterns &
structures
Example Algorithms Decision Trees, SVM,
Neural Networks
k-Means, DBSCAN,
Apriori
Applications Spam Detection, Disease
Prediction
Market Segmentation,
Anomaly Detection
8. Example Illustrations
• Supervised Learning: Predicting house prices
based on past sales (Regression)
• Unsupervised Learning: Grouping customers
into similar categories based on spending habits
(Clustering)
• Visual Representation: Graph showing
classification boundary vs. clusters
9. Common Applications of Supervised Learning
• Healthcare: Diagnosing diseases based on
medical records
• Finance: Credit risk assessment
• Natural Language Processing: Sentiment
analysis
• Computer Vision: Facial recognition
10. Common Applications of Unsupervised Learning
• Market Segmentation: Identifying different
customer groups
• Anomaly Detection: Detecting fraud in
transactions
• Recommender Systems: Grouping similar users
for personalized recommendations
11. Case Study – Supervised Learning (Spam Email
Detection)
• Problem Statement: Classify emails as spam or
not spam
• Dataset: Features include sender, subject, email
text
• Approach: Use Naïve Bayes classifier
• Outcome: Model achieves 95% accuracy in
detecting spam
12. Case Study – Unsupervised Learning (Customer
Segmentation)
• Problem Statement: Identify customer groups
for targeted marketing
• Dataset: Includes purchase history, browsing
behavior, demographics
• Approach: Apply k-Means clustering to group
similar customers
• Outcome: Business personalizes promotions
leading to increased sales
